EVALUASI BERBASIS SIMULASI SKRINING KEKERASAN PSIKOLOGIS DAN TRAUMA KETIDAKSETIAAN PASANGAN PADA IBU HAMIL

Abstract

Trauma and psychological violence related to partner infidelity often remain undisclosed during routine antenatal care, particularly when face-to-face interviews are perceived as unsafe or uncomfortable. This study aims to evaluate the initial measurement performance of an adaptive digital screening model and its detection rates. This study combines a descriptive analysis of an initial 80-case screening dataset with fully simulation-based item-level and paired pre-post datasets. The model assesses psychological violence, trauma from partner infidelity, anxiety, depression, social support, disaster-related vulnerability, and safety red flags. Simulated changes in detection performance were analyzed using the t-test, Wilcoxon signed-rank test, McNemar test, and Cohen’s dz. Findings revealed moderate-to-high psychological violence in 55% of cases and moderate-to-severe trauma in 60% of cases. All 25 simulated items met validity and reliability criteria. Detection performance scores significantly increased from 54.62 to 70.96 (p < 0.001; dz = 2.648). Detected cases increased from 38 to 65 cases (a 71.05% increase), accompanied by significant shifts in paired pre-post detection status. The adaptive model demonstrates methodological feasibility for privacy-preserving and risk-stratified antenatal screening. Nevertheless, field validation against direct clinical assessments remains necessary prior to routine implementation in clinical practice.